Your Dashboard Is Gaslighting You: How to Stop Trusting Vanity Metrics and Start Reading Real Data

There's a particular kind of confidence that comes from opening Google Analytics, glancing at a rising line graph, and thinking: brilliant, we're doing well. It feels productive. It feels data-driven. And in many cases — particularly for small and medium businesses across Surrey and the wider UK — it is almost completely misleading.

We've sat across the table from enough business owners in Dorking, Guildford, and Reigate to know that this isn't a niche problem. It's endemic. The tools are powerful, the interfaces are polished, and the numbers look authoritative. But authority and accuracy are not the same thing.

Let's talk about what your analytics are actually telling you — and, more importantly, what they're not.

The Bounce Rate Trap

Bounce rate is probably the most misunderstood metric in web analytics. The common assumption: a high bounce rate means people are leaving your site immediately because they hate it. The reality: it's far more nuanced, and acting on that assumption can cause genuine harm.

Consider a local accountancy firm in Leatherhead we worked with a couple of years back. Their homepage bounce rate was sitting at around 78%, and their previous agency had flagged it as a critical problem. The proposed fix? A complete homepage redesign, new copy, new layout — a four-figure spend.

Before touching a single pixel, we dug deeper. It turned out that a significant chunk of that 78% was made up of returning visitors who were landing directly on the contact page via a saved bookmark, getting the phone number, and leaving. They weren't bouncing in frustration. They were bouncing in satisfaction. The page was doing exactly what it needed to do.

The redesign would have solved a problem that didn't exist.

Sessions, Users, and the Counting Problem

Here's another one that trips people up: session counts and user counts are not interchangeable, and conflating them can seriously distort your picture of audience growth.

A session is a single visit. A user is a person (or at least, a device). One user can generate dozens of sessions. If your development team or your own staff are regularly accessing the site — testing pages, checking content, reviewing updates — they're inflating your session numbers without representing a single genuine prospective customer.

We've seen this skew reporting dramatically, especially for smaller sites where internal traffic represents a meaningful percentage of total visits. The fix is simple: filter out known internal IP addresses in your analytics configuration. But it's a step that's overlooked with remarkable frequency.

Once that filter is in place, some businesses discover their 'growing audience' is actually flat. That's uncomfortable information — but it's honest information, and honest information is what good decisions are built on.

Average Time on Page: The Metric That Lies Both Ways

Long time on page sounds positive. Surely that means people are reading, engaging, absorbing your content? Sometimes. But a high average time on page can also mean your users are confused, struggling to find what they need, or simply walked away from their computer mid-session and left the tab open.

Conversely, a low time on page isn't always bad. If someone lands on your pricing page, spends 22 seconds, and then clicks through to your contact form — that's a conversion funnel working exactly as intended.

A retail business near Horsham came to us frustrated that their product pages were showing low engagement times. They'd invested in lengthy product descriptions and assumed customers weren't reading them. In reality, a heatmap analysis revealed that users were scrolling quickly to the price and the buy button — the descriptions were fine, the page was just well-optimised for decisive buyers. The real issue was something the time-on-page metric couldn't reveal at all: a checkout flow that was losing people at the payment stage.

That's the thing about vanity metrics. They don't just mislead — they actively distract from the real problems.

What You Should Actually Be Measuring

So what does useful analytics work look like? Here's how we approach it:

Start with goals, not graphs. Before you open your dashboard, write down the two or three things your website actually needs to do. Generate enquiries? Sell products? Get newsletter sign-ups? Every metric you look at should be evaluated against those goals.

Use event tracking deliberately. GA4 in particular is built around event-based data, but events only tell you something useful if you've defined them clearly. Track form submissions, button clicks, file downloads, video plays — the actions that indicate genuine intent.

Segment relentlessly. Overall traffic numbers are almost meaningless. Break your data down by source, device type, landing page, and geography. A campaign that's performing brilliantly among mobile users in Surrey might be completely invisible to desktop users in the Midlands.

Cross-reference with real feedback. Analytics tell you what is happening. They rarely tell you why. Combine your data with user surveys, session recordings, and direct customer conversations. The picture that emerges is almost always more complicated — and more actionable — than the dashboard alone suggests.

The Dork's Approach to Data

We like to think of good analytics work as a bit like proper research — methodical, sceptical, and resistant to easy conclusions. At WebDorking, we've built a habit of questioning the first interpretation of any dataset before acting on it. The numbers that seem most obvious are often the ones worth scrutinising hardest.

Your website is one of your most significant business assets. The data it generates should be informing your investment decisions, your content strategy, and your development priorities. But only if you're reading it correctly.

If your current analytics setup is giving you a dashboard full of rising lines and reassuring percentages — but your enquiry volumes aren't matching that optimism — it might be time to look a little closer. The answers are usually in there. They're just not always where you'd expect to find them.